SHORT RANGE ORDER OF METALLIC GLASSES

(1)

HAL Id: jpa-00220370

https://hal.archives-ouvertes.fr/jpa-00220370

Submitted on 1 Jan 1980

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of

sci-entific research documents, whether they are

pub-lished or not. The documents may come from

teaching and research institutions in France or

abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est

destinée au dépôt et à la diffusion de documents

scientifiques de niveau recherche, publiés ou non,

émanant des établissements d’enseignement et de

recherche français ou étrangers, des laboratoires

publics ou privés.

SHORT RANGE ORDER OF METALLIC GLASSES

A. Schaafsma, I. Vincze, F. van der Woude, T. Kemèny, A. Lovas

To cite this version:

A. Schaafsma, I. Vincze, F. van der Woude, T. Kemèny, A. Lovas.

SHORT RANGE ORDER

(2)

JOURNAL DE PHYSIQUE _{ColZoque C8, suppZ&rrent}au noE, Tome 42, aoiit 2980, page

C8-246

SHORT RANGE ORDER OF METALLIC GLASSES

++

A.S. Schaafsma, I. ~incze+, F. Van der Lloude. T. Kemeny _{and A. Lovas}₊₊

S o l i d S t a t e Physics Laboratory,Material.s Science Center, l k t i v e r s i t y of Groningen, Melkweg

;,

*,

9718 EP Groningen, The Netherlands.

Central Reseurch I n s f Z t u t e for Phgsics, Budayest, fiungam~.

Abstract.

-

_{Present PGssbauer investigations of melt-quenched (Fe Xi}

) 5B (0

5

x

5

0.75) glasses show a change in the geometrical arrangement of near-neighbours c i ~ z e l ~ ?ol?owine, that of the crystalline counterparts (Fe B is tetragonal, Ni B is orthorhombic). The chemical short-range order of sputte-

3

red amorphous re _{1-x x}B (0.09

5

x

5

0.30) alloys characterized via the Fe hyperfine field distribution is very similar to th2l of liquid quenched samples.

The physical properties of amorphous alloys are determined by the short-range atomic arrangement which is not well known due to the lack of direct experimental information. Msssbauer spectroscopy is very useful in the study of chemical short-range order because of its sensitivity to nearest-neigh- bour environments. In the following we will res- trict our discussion for ferromagnetic transition metal-metalloid glasses. Two types of information can be obtained from Mossbauer experiments: I , in these systems the iron hyperline field is proportional [l] to the iron magnetic moment which is determined [2] mainlj by the number of nearest metalloid neighbours and quite insensitive [3] for the transition metal neighbourhood (in the case of Fe, Co or Ni) and for the actual geometrical arrange- ments of the metalloids. Thus the hyperfine field distribution is determined by the metalloid distribution around the iron atoms (it provides the probability of an iron atom with a given number of metalloid neighbours). The isomer shift gives similar information but is less sensitive. 2, the geometrical arrangement of metalloid (and transition metal) neighbours is reflected in the quadrupole interac- tion A 6 which is proportional to the electric

Q

field gradient at the nuclear site and vanishes for a cubic environment. This is less sensitive for the number of metalloid neighbours but strongly influ- enced by their actual geometrical symmetry.

In the following results will be presented which show: 1, a change in the chemical short-range order (CSRO) of liquid quenched (Fel-xNix)75B25 glasses

'on leave from thercentral Research Institute for Physics, Budapest.

for large Ni concentrations, and 2, a similarity of the CSRO of sputtered and liquid quenched Fel-xBx glasses.

It has been shown [4] that in the liquid quenched Fel-xBx (0.15

2

x

2

0.25) glasses the Fe0.75B0.25 composition plays a unique role in the sense that its CSRO can be well approximated by that of the metastable tetragonal Fe3B compound formed during the crystallization of these glasses. This structure is isostructural to Fe P and has three crys-

3

tallographically inequivalent iron sites in equal number with 28, 3B and 4B nearest nelghbours. For the substitution of Ni the tetragonal structure transforms into the orthorhombic structure of Ni B.

3 This cementite type of crystal contains two crys- tallographically inequivalent Ni sites with 2B and 38 nearest neighbours in 1:2 ratio. The difference in the symmetry of the environments (and also the stronger hybridization due to the shorter TM-M dis- tances) is clearly reflected in the paramagnetic Mossbauer spectra of the tetragonal Fe3B and orthorhombic (Fe0.33Ni0.67)3B (Fig. la and lb). The

-

average quadrupole splittings AE extrapolated to

Q

800 K are 0.40 mm/s and 0.70 mm/s, respectively.

It has been shown [ 5 ] by the Massbauer investiga- tion of liquid quenched (Fel-xNix),5B25 glasses that both the iron hyperfine field and the Curie temperature closely follow those of the crystalline counterparts

-

thus indicating a similar local structure and also a change with increasing Ni content in the CSRO of these glasses. The paramagnetic spectra shown in Fig. lc and Id support this con- clusion. (Sincethe amorphous Fe 0 . 7 5 B 0.25 crystalli- zes before the Curie point the spectrum of amorphous Fe B is shown but

hE

was found to be

0.84 0.16

Q

(3)

FIG. 1. Paramagnetic Mossbauer spectra of crystalline tetragonal Fe B (a), orthorhombic

3

(Fe0.33Ni0.67)3B (b), amorphous Fe0.84B0.16 (C) and (FeO. 33Ni0.67)75B25 (d) '

-

quite independent of the B content). AE (600 K) =

Q

0.40 mm/s and 0.57 mm/s for a Fe0.84B0.16 and a (Pe0.33Ni0.67),5B25, respectively which show the same trend observed for the crystalline alloys. Thus these data show not only that the CSRO of the amorphous and the crystalline structure at this special composition is similar but also that it has changed for the Ni substitution. This observation is beyond the scope of the different dense random

packin& (D9P) models where no change is expected because of the similar radii of Fe and Ni atoms. Also the comparison of the CSRO of liquid quenched

(I&) and sputtered (SP) Fe-B glasses via their Moss- bauer data does not support models based on DRP of spheres. A priori it is not expected that the CSRO will be similar in both cases. For example, the pre- sence of deep eutectics in the glass forming concentration range suggests that strong interactions in the liquid play a significant role in glass forma- tion [6]. These interactions are obviously absent in the gas phase which might result in a chemically more random local structure for sputtered glasses. Fel-xBx (0.09

5

x - < 0.32) glasses,were prepared by argon DC getter sputtering onto a liquid nitrogen cooled substrate. The sputtering apparatus was es- sentially the one described by Theuerer and Hauser

[7], but adapted in order to carry out in situ Moss- bauer measurements. The substrate material was boron nitride or aluminum foil, both giving the same results. After warming up to room temperature the sputteres layers (thickness 5 to 1 0 microns) were still X-ray amorphous.

The tf8ssbauer spectra for liquid quenched and sputtered Fe.85B.15 are shown in Fig. 2. The most stri-

(4)

C8-248 JOURNAL DE PHYSIQUE

king feature is the difference in the relative in- tensity of lines 2 and 5, which means that for the I& ribbons the magnetization direction is preferen- tially in the plane as determined by the shape anisotropy while for the SP films it is perpendicular. The strong perpendicular anisotropy of the SP films

is probably caused by a columnar microstructure which was found to be characteristic for a wide va- riety of low temperature vapour quenched films 181. Apart from this difference in anisotropy the ?.36as- bauer spectra of LQ and SP FeB alloys are identical within the error in the whole concentration range. This is shown in fig. 3 where the mean hyperfine

FIG. 3. Mean hyperfine field, < H >, and standard deviation, a, of hyperfine field distribution at T

= 78 K for sputtered (open circles) and liquid quenched (closed circles) FeB glasses.

field < H > = /H.P(H) dH and the standard deviation of the hyperfine field distribution P(H) are given as a function of B concentration. In both cases P(H) is rather symmetric and the width is independent of concentration, whereas the mean hyperfine field decreases with increasing 9-concentration following the same trend as shown by the crystalline a-Fe, Fe3B (tetragonal) and Fe2B phases Fig. 4 shows the P(H) of sputtered Fe0.90B0.10 and Fe 0. 70B0. compared with calculations based [2] on different models. The first model uses the concept of DRP of hard spheres, the second one emphasizes the type of SRO occurring in Fe3B (quasi-crystalline model, QC). In both cases the P(H) is calculated from the probability distribution of the number of 9-atoms around an Pe atom p(nB) having the dis-

FIG. 4. Experimental hyperfine field distributions (full curves) compared with calculated ones based on a DRP model (dashed curve) and a quasi-crystalline model (dotted curve), as explained in the text. The vertical bars indicate positions and relative intensities of the individual components in the DRP model. The inserts show the distribution functions of B nearest-neighbours around an Fe atom in both models.

Crete H(nB) valuesfound incrystallineFe-Bphases and subsequent Gaussian broadeningwith standarddeviation AH/H = 0.105 determined for a Fe.75B.25. The p(ng) distribution of the DRP model was that of Jansen et al. 191 for a radius ratio r /r = 0.50. In t h e Q C

B Fe

model (where the Fe.,5B.25 composition is stoichio- metric) a random substitution of Fe onto B sites (or B onto Fe sites) was assumed. The QC model gi-

ves a better agreement though at 1 0 at% B it overestimates the low B coordinations (nB = 0,l) and at 3 0 at% it overestimates the high B coordinations (nB = 5,6). This means that the assumed random substitution of F e by B (or B by Fe) is not realistic. There is a tendency for preferred substitution pre- serving the type of local surrounding of Fe with B present in crystalline Fe3B (nB = 2, 3 , 4) and FeZB (n, = 4 ) .

(5)

t h e Advancement of Pure Research (ZWO).

REFERENCES

[I]

K . R a j , J . Durand, J . I . Budnick and S. S k a l s k i ,

J . Appl. Phys.

2,

1676 (1978); F. van d e r Wou-

de and I . Vincze, J . de Physique

41,

C1-151 (1980).

[2] I . Vincze, D.S. Boudreaux and M . Tegze, Phys. Rev. B 19, 4896 (1979).

[3] I . Vincze, F. van d e r Woude and J . Balogh, J .

d e Physique

41,

C1-257 (1980).

[4] I . Vincze, T. Kembny and S. A r a j s , Phys. Rev. B 21, 937 (1980).

[5] I . Vincze, F. van d e r Woude, T. Kemhny and A.S. Schaafsma, J . Magn. IAagn. Mat.

15-18,

1336 (1980).

[6] J . J . Gilman, P h i l . Mag.

e,

577 (1978). [7] H . C . Theuerer and J . J . Hauser, Trans. Metallur-

g i c a l Soc. of AIME

233,

588 (1965).

[8] A . G . D i r k s and H . J . Leamy, Thin S o l i d Films

47,

219 (1977).